Digital Image Steganography Analysis And Evaluation System

With the rapid development of Internet and information security technology, steganography, as an important kind of information security technology, does obtain so much attention from researchers. Meanwhile, steganalysis has become the hot topic in the field in recent years. With continuous occurrence of more and more steganalytic algorithms, how to compare and evaluate those algorithm becomes antoher important question.In this dissertation, a thorough and complete review on the related theories and technical foundations of steganalysis is made, the index system for evaluating steganalytic algorithms is established, and the software tool running on Windows is designed and developed. The contribution of the thesis is listed as follows.(1) A summary is made on main steganalytic algorithms for digital image that are commonly classified into specific estimation class, specific detection class, and universal detection class. Then the basic principles for typical algorithms of each class are introduced in detail.(2) With the considartion of embedding rates and testing images characteristics, four-level index evaluation system is constructed for different steganalytic algorithms mainly based on their detection performance and speed.(3) On the basis of multi-level index evaluation system, a new approach for assessing steganalytic algorithms based on AHP and entropy weight fuzzy comprehensive evaluation method is proposed by introcuding entropy weight method to modify weight coefficients in traditional fuzzy comprehensive evaluation method.(4) The evaluation system for steganalysis are designed and implemented on Windows platform. Representative algorithms of each of the three classes of steganalytic algorithms mentioned above are tested, and the testing results are also analyzed and discussed.At last, a conclusion is drawn on the shortcomings of the dissertation, and the future research contents on this topic are prospected.